WORKING P A P E R. Veterans Health Administration Mental Health Program Evaluation Technical Manual

(1)

Veterans Health

Administration Mental

Health Program Evaluation

Technical Manual

MARCELA HOROVITZ-LENNON,

KATHERINE E.WATKINS, HAROLD ALAN PINCUS,

LISA R. SHUGARMAN, BRAD SMITH,

TERYN MATTOX, THOMAS E. MANNLE, JR.

WR-682-VHA February 2009

Prepared for the Veteran’s Health Administration

W O R K I N G

P A P E R

This product is part of the RAND Health working paper series. RAND working papers are intended to share researchers’ latest findings and to solicit informal peer review. They have been approved for circulation by RAND Health but have not been formally edited or peer reviewed. Unless otherwise indicated, working papers can be quoted and cited without permission of the author, provided the source is clearly referred to as a working paper. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.

(2)

(3)

1

Preface

The Veterans Health Administration (VHA) in the U.S. Department of Veterans Affairs (VA) provides mental health and medical treatment for veterans. In order to deliver top-quality medical care to veterans, VA must ensure that program goals are being met and that the services available are adequate. To this end, the VA Office of Policy and Planning contracted with Altarum Institute and the RAND–University of Pittsburgh Health Institute (RUPHI) to conduct an independent study to evaluate its mental health programs. This evaluation is mandated by Congress under the

Government Results and Performance Act of 1993 and under Title 38 of the U.S. Code (Veteran’s Benefits). The results of this study will be used to inform VHA policy and operational decisions for mental health.

Critical to this evaluation was the development of a comprehensive set of mental health performance indicators based upon available VHA administrative and medical record data. The team applied the Institute of Medicine quality of care framework in the identification of performance measures. The Institute of Medicine has defined six quality domains: effectiveness, efficiency, equitability, safety, and patient centered care. i

Furthermore, VHA identified critical domains of quality within its own organization, including: diagnosis and assessment, treatment, chronic disease management and rehabilitation.

This report presents the technical specifications for the performance indicators developed during the study. The project team drew upon existing performance

indicators developed for the mental health population, clinical practice guidelines for mental health diagnoses, and the clinical expertise of team members and advisors in the development of the indicators. The strength of evidence for each indicator was assigned based upon guidelines from the Agency for Healthcare Research and Quality, as

adopted by VHA.

The VHA has contributed directly to the development of the quality indicators described in this technical manual through an advisory group composed of

representatives from the VHA Patient Care Services, the VHA Office of Mental Health, several field practitioners, and contractors. This advisory group collaborates with the evaluation team through input on the evaluation’s scope and methodologies.

The contents of this report will be of interest to policymakers, health care organizations, and clinical practitioners who are engaged in activities related to the improvement of mental healthcare quality.

(4)

(5)

3

PART II: ADMINISTRATIVE DATA INDICATORS ... 28

SCHIZOPHRENIA ... 29

BIPOLAR I DISORDER ... 41

POSTTRAUMATIC STRESS DISORDER ... 52

MAJOR DEPRESSIVE DISORDER ... 54

SUBSTANCE USE DISORDER... 61

CROSS-CUTTING INDICATORS ... 73

Part III: Medical Records Review Indicators ... 117

SCHIZOPHRENIA ... 119

BIPOLAR DISORDER... 132

POST TRAUMATIC STRESS DISORDER (PTSD) ... 144

MAJOR DEPRESSIVE DISORDER (MDD) ... 160

SUBSTANCE USE DISORDER (SUD)... 173

CO-OCCURRING DISORDERS ... 200

CROSS-CUTTING: PSYCHOSOCIAL NEEDS INDICATORS... 211

CROSS-CUTTING: SUICIDALITY INDICATORS... 227

CROSS-CUTTING INDICATORS ... 231

(6)

4

SUMMARY

Questions are being raised nationally about access to and quality of mental health care, both within the Veterans Health Administration (VHA) and in the public and private sectors more broadly. Numerous studies have documented the discrepancies between mental health care that is known to be effective and mental health care that is actually delivered.ii These gaps are important because mental health conditions are a leading cause of disability and death and have serious economic, social, and personal consequences.iii_{Concurrently, U.S. policy makers}

and medical professionals are increasingly recognize that quality mental health care can lead to better, healthier lives for those with mental illness, and that performance measurement plays a key role in improving health care delivery and, ultimately, patient outcomes.

In 2006, the U.S. Department of Veterans Affairs (VA) funded an independent study to evaluate the quality of mental health services provided to veterans. This study is mandated by the Government Results and Performance Act of 1993, which requires federal agencies to independently evaluate high-volume, high-cost programs, and Title 38 of the U.S. Code, which regulates Veteran’s Benefits. It represents one of the largest and most comprehensive

evaluations of mental health services ever conducted. The evaluation focuses on five high-volume, high-cost mental health diagnoses that have the potential to greatly impair quality of life for veterans:

Schizophrenia Bipolar disorder

Post-traumatic stress disorder (PTSD) Major depressive disorder

Substance use disorder.

This evaluation of the VHA mental health services is designed to present new

information about how well VA is translating the promise of improved mental health care into better, healthier lives for veterans. In particular, the study team will examine whether specific gaps in services identified as targeted areas of improvement in the MHSP have been reduced by the implementation of the plan, and whether investments and/or other enhancements in VA mental health and substance use treatment services under the plan have had a positive impact on capacity, utilization, staffing, and individual users over the study period.

In order to develop and select measures that would be viewed as meaningful and useful in addition to valid, the team’s work was guided by policy documents that identified the aims and characteristics of high quality health care. The Institute of Medicine’s quality of care paradigm was used explicitly to categorize all potential measures and to ensure indicators covered all six domains of effectiveness, safety, efficiency, timeliness, patient-centeredness and equity. ivv The VHA Mental Health Strategic Plan was modeled after the Report by the President’s New

Freedom Commission on Mental Health; both identified the consumer and family as the drivers of mental health care, focusing attention on the concept of recovery and on the elimination of disparities in the availability and quality of mental health services. Together these documents provided the social and political context for development and selection of the measures herein. Below we have documented the methodology employed in the development of mental health indicators.

Indicator development process

1. Conduct a Systematic Search for Previously Identified, Grade I Performance Indicators.

We conducted a systematic review of the literature including studies, technical reports, reviews, electronic databases, etc., manual review of relevant bibliographies, and outreach to

(7)

5

experts and industry representatives to identify an exhaustive pool of relevant performance indicators that were either in the public domain or were being prepared for near-term

dissemination. All relevant measures were retrieved and the team reviewed the methodology used in their design to assess their quality. We abstracted each performance indicator, noting its data source, the disorder to which it applied the strength of the evidence for the process measured by the indicator, and IOM domain.

2. Identify recommendations with empirical support that are not covered by the existing measures, and create new performance indicators to address these gaps.

We reviewed VA and APA Clinical Practice Guidelines for the 5 disorders included in the program evaluation (the VA CPG for psychoses includes recommendations for both

schizophrenia and bipolar disorder), and listed all individual recommendation statements. Multi-part recommendations were separated into individual parts and duplicative

recommendations were deleted. We defined key terms, examined the recommendations for inconsistency or ambiguity, and produced a list of explicit, unambiguous measures that had empirical support for the process-outcome link. Where discrepancies existed between the APA and VA guidelines the team consulted outside experts and discussed the discrepancy until consensus was reached.

3. Select measures for further technical specification.

Because of the large number of candidate measures, we engaged in a systematic selection process. First, we identified whether the data needed to populate the indicators existed in the necessary form in either the administrative or in the medical record, and recommendations that could not be defined operationally because of lack of data were

eliminated. Next, the research team reviewed the measures for meaningfulness and feasibility, and described the measures’ predictive validity through an evaluation of the strength of the process-outcome link. A subset of measures was selected to be reviewed by external clinical experts who further pruned them on the basis of clinical significance. All measures were reviewed with a VA clinical advisory group, whose members were selected for their clinical expertise and familiarity with the subject matter. The advisory group evaluated

recommendations for validity and feasibility, and usefulness for VA’s operational management and strategic leadership. Lastly, VA and VHA leadership rated the indicators on their

importance to the VHA and contribution to presenting a comprehensive quality profile. As a result of this process, we identified a core set of measures that were valid, feasible, and a VA priority. Most of them described processes that were identified with acute treatment.

4. Generate a new set of measures pertaining to the psychosocial aspects of care.

Because the process used above required measures to have an empirical basis of support, the domains of patient-centeredness and recovery were neglected. Although not evidence-based or guideline-supported, both domains are endorsed by the Institute of Medicine and the VA Mental Health Strategic Plan as critical to quality. We therefore used a collaborative process between the research team and the VA clinical advisory group to identify key constructs pertaining to patient-centeredness and recovery. Among the many possible constructs, we chose to focus on the psychosocial aspects of care such as attention to social supports, housing and employment. Indicator development involved recruiting experts and engaging them in the process of identifying a core set of cross-cutting psychosocial indicators. Because of the difficulty evaluating the predictive validity of the psychosocial aspects of care, they will be used descriptively.

(8)

6

5. Develop technical specifications for finalized indicators and categorize their strength of evidence

We generated detailed technical specifications for all finalized performance indicators with respect to VHA administrative data and electronic medical records, and identified data sources that efficiently provided the information necessary to populate the indicators. Each indicator contained an indicator statement and executive summary describing the source(s) of the specifications and clinical rationale for the selected indicator. We also included the indicator grade, which reflected the strength of the process-outcome link, and whether the indicator would be used as a benchmark or descriptively. We created numerators and denominators for each indicator based on the data that would be available, and defined the population to which the indicator applied. For example, if the indicator applied only to people in a new treatment episode, we defined the term ‘new treatment episode’. All clinical and

measurement terms were defined operationally, and we summarized anticipated data collection problems and other feasibility issues. These included any problems that we could foresee prior to starting abstraction, such as data elements that might be time-consuming to collect or which required a judgement to be made by the abstractor. For complex processes of care with multiple components of varying clinical or outcome relevance (e.g., delivery of CBT/SST or assessment of mental status), we sought expert input to select and operationalize critical components. Technical specifications were reviewed by both external clinical experts and the VA Advisory group in order to ensure that specifications were both feasible given the data available, and meaningful to this particular population.

We categorized indicators according to the strength of the process-outcome link using the grading system developed by the AHRQ’s US Preventive Services Task Force.vi_{Grade I}

measures are those where the link between process and outcome has been established through randomized clinical trials, grade II measures are supported by well-designed, non-randomized trials, and grade III measures are supported by expert opinion. A caveat to drawing conclusions from this grading system is that sometimes the outcomes literature may not be specific enough about the ingredients of the intervention that are critical to its

efficacy/effectiveness. For example, although randomized controlled trials have established the value of psychotherapy in the treatment of several disorders, not enough evidence exists to ascertain the minimum “dose” (or number of sessions) and duration required for the outcome advantage to emerge. We also note that the grading does not reflect translational validity, or the certainty that the technical specifications accurately reflect the process of care they are trying to capture.

6. Determine data abstraction elements and sequence of abstraction

Starting with the technical specifications developed above, we described the data abstraction elements and abstraction sequence for each indicator. Since many indicators required overlapping information, we removed redundancy and grouped questions for efficiency. For example, all questions about medications were placed together,, since the medications prescribed to a veteran are found in a single section of the record. We created abstraction forms for each diagnosis.

7. Pilot test indicators for translational validity and performance

Clinical nurse abstractors piloted each indicator for timing and performance using pencil and paper and modifications were made in order to keep data collection time to a minimum. We

(9)

7

found that some data elements were not found in the part of the medical record to which we had access, and, after review with the clinical advisory group, deleted these indicators. After the initial paper and pencil pilot test , an electronic abstraction form was created and a second pilot test was performed to make sure that the questions flowed correctly and that there were no programming errors.

Discussion

In this report we present a comprehensive set of indicators for evaluating the

performance of mental health care systems with two different data sources, administrative and medical records. One of the greatest difficulties in evaluating mental health care is obtaining meaningful data to measure the key elements of the system. In order to evaluate the structure of care, we developed indicators that used a combination of both data sources available, while recognizing that both sources of data, either singly or in combination, have inherent strengths and weaknesses.

The main strength of using administrative data is their availability and comprehensive enumeration of the study population.vii_{Moreover, the databases were relatively large, enabling}

the study team to analyze population subgroups and specific geographic areas separately, which was particularly useful, since most problems related to access and availability are not uniform across populations or within areas. In many cases, however, items were missing or the accuracy of the information provided could not be guaranteed. This is not uncommon when data are collected and used for different purposes. Other studies also support the use of

administrative data combined with chart review to assess performance.viii

While the structure-process-outcomes evaluation model presented herein holds promise for advancing the science of mental health care quality improvement both within and outside the VHA, a few final caveats are in order.

First, in any health care system, the progression from evidence-based practice

guidelines to performance indicators to improved patient outcomes is fraught with complexity. Great care must be taken to measure precisely what is intended to be measured through effective and efficient documentation so that the burden of measurement does not outpace quality care provision. In addition, continued awareness of the complicated linkages between evidence-based practice and individual patient preferences and outcomes is essential. As recent studies amply demonstrate, even the most basic of evidence-based practice

improvements can result in different outcomes for different patients and for different reasons. Attention must also be paid to ensuring that quality improvement becomes a part of the fabric of care at both the organizational and individual levels, through resource investment, staff training, etc.

Second, not all mental health care systems look or operate like the VHA mental health care system. Public and private sector mental health care functions largely as a cottage industry, with the majority of psychiatrists practicing in solo or two-physician practices; information technology is less well developed; there are few centralized administrative databases; and there is no single entity or organization responsible for implementing and monitoring quality improvement strategies. While these differences must be recognized and addressed in the context of ongoing quality improvement, the same high quality standards should nevertheless apply.

Third, to what extent this model can be adapted for use in other systems and in other contexts is not clear. It is possible that certain components of the model will be more suitable for mental health quality improvement efforts at the national or state levels or in large systems (e.g., managed care networks), while others will work well in more localized contexts (e.g., community mental health centers).

(10)

8

VA has undertaken the most extensive, systematic, and rigorous evaluation of the mental health care delivery ever conducted. Although this quality improvement effort is still in its early stages, and much remains to be learned, the framework, methodology, and preliminary results offer a fertile ground upon which other stakeholders in the mental health field can continue to build and expand both in the near- and longer-term.

(11)

9

INTRODUCTION

This technical manual is presented in three main parts. Part I defines the key terms used in the description of indicators, part II describes the administrative data indicators, and part III describes the medical record indicators.

Part I, the Key Definitions Document (KDD) defines the relevant populations, types of treatment encounters, treatment episodes for each of six main diagnoses (bipolar disorder, schizophrenia, major depression disorder (MDD), post-traumatic stress disorder (PTSD), substance use disorder (SUD), and co-occurring disorders), and additional concepts such as specialty mental health, licensed mental health provider, licensed mental health prescribing provider, licensed prescribing provider, and psychotherapy. This document is a companion to support the use of any of the indicators in parts II and III.

Part II describes 31 indicators designed for administrative data analysis including indicators specific to the treatment of bipolar disorder, schizophrenia, substance use disorder, major depressive disorder, and post-traumatic stress disorder as well as cross-cutting indicators that apply to two or more of the diagnoses considered.

Part III describes 57 hybrid indicators that integrate data from administrative and medical records. In addition to documenting indicators for each of the six main diagnoses and cross-cutting indicators this section also includes suicide indicators to review assessment for suicide ideation and follow-ups for suicidal patients. And psychosocial indicators to evaluate whether mental health patients receive mental status exams appropriate to their diagnosis and

psychosocial assessments and support across the domains of housing, social support, and employment.

Strength of evidence:

The Altarum/RUPHI team has adopted the same grading system as the VHA in its Clinical Practice Guidelines, described in the following table:

Strength of Evidence Grade Description

I Evidence is obtained from at least one properly randomized controlled trial (RCT).

II Evidence is obtained from well-designed cohort, case-controlled, case-controlled, or time series trials without randomization.

III Opinions of respected authorities are based on clinical experience, descriptive studies and case reports, or reports of expert committees.

We used the strength of evidence linking the process of care to desired outcomes to define whether the indicator would be used for benchmark or descriptive purposes. Benchmark indicators are those supported by grade I evidence for which data were available and could be collected. The remaining indicators are used for descriptive purposes only. Of the 88 indicators developed 21 are benchmark indicators and the remaining 67 are descriptive.

(12)

10

Study Period:

In several indicators, we use the term “study period,” which in this analysis includes fiscal year (FY) 2007. In certain instances we have extended the scope of analysis to include data from FY 2008 or a look-back period into FY 2006. In those cases, the change is noted in the text of the indicator.

Data Sources:

For all indicators, we are relying on administrative and medical record data to define the numerator and denominator. Indicators that require pharmacy administrative data may be operationalized on a smaller sample unless we gain approval to receive the complete pharmacy file for all patients in our universe as defined in the Key Definitions Document.

Industry standard indicators:

Where possible we have used indicators that have been cited by major mental health care performance indicator clearinghouses. These indicators have been previously developed and substantiated with evidence or clinical consensus. We will cite these clearinghouses in the rationale statements for many performance indicators in this document. Below we have

included a brief summary of each of these clearinghouses from their own documentation. Center for Quality Assessment and Improvement in Mental Health (CQAIMH)

CQAIMH inventory of quality measures provides a searchable database of more than 300 process measures for assessment and improvement of mental health and substance abuse care. These measures have been developed by government agencies, researchers,

professional organizations, accreditors, health systems, employer purchasers, consumer coalitions, and commercial vendors. Each measure is accompanied by a clinical rationale, numerator and denominator specifications and information on data source, domain of quality, evidence basis and developer. The inventory can be searched by these characteristics as well as by diagnosis, demographics, type of treatment and clinical setting. Funding for the Inventory was provided by the Agency for Healthcare Research and Quality (AHRQ), Substance Abuse and Mental Health Services Administration (SAMHSA), and the Evaluation Center@HSRI. Source: http://cqaimh.org/

STABLE: Standards for Bipolar Excellence

The STABLE Performance Measures & Toolkit provides rigorously developed tools for quality assessment and improvement of care for bipolar disorder. Fifteen performance

measures were developed on the basis of research evidence, expert consensus and formal testing of reliability and validity. Detailed specifications, medical record abstraction forms and performance results from 80 outpatient sites are provided. The toolkit provides numerous resources to improve performance in clinical practice including instruments to screen for depression and mania as well as to monitor symptoms and functioning over time. In addition, there are tools to assess for suicide risk, co-morbid substance use and medication side effects; to provide patient education; and to assist with diagnostic coding. STABLE was led by a

National Coordinating Council of bipolar and measurement experts as well as leaders of national professional associations. Funding was provided by AstraZeneca LLP.

Source: http://www.cqaimh.org/stable.html National Quality Forum (NQF)

The National Quality Forum (NQF) is a not-for-profit membership organization created to develop and implement a national strategy for health care quality measurement and reporting. A shared sense of urgency about the impact of health care quality on patient outcomes, workforce

(13)

11

productivity, and health care costs prompted leaders in the public and private sectors to create the NQF as a mechanism to bring about national change.

Source: http://www.qualityforum.org/

Healthcare Effectiveness Data and Information Set (HEDIS) and National Committee for Quality Assurance (NCQA)

Developed and maintained by NCQA, HEDIS is a tool used by more than 90 percent of America's health plans to measure performance on important dimensions of care and service. Altogether, HEDIS consists of 71 measures across 8 domains of care. Because so many plans collect HEDIS data, and because the measures are so specifically defined, HEDIS makes it possible to compare the performance of health plans on an "apples-to-apples" basis. Health plans also use HEDIS results themselves to see where they need to focus their improvement efforts.

To ensure that HEDIS stays current, NCQA has established a process to evolve the measurement set each year. NCQA’s Committee on Performance Measurement, a broad-based group representing employers, consumers, health plans and others, debates and decides collectively on the content of HEDIS. This group determines what HEDIS measures are included and field tests determine how it gets measured.

(14)

12

(15)

13

Defining Study-Relevant Populations

Ia. Cohort diagnoses

In the following table are ICD-9 codes used to define each of our cohorts. To be included in the study population, clients must have at least two outpatient encounters on different days or one inpatient episode during the study period for any reason. Patients are assigned to a diagnostic cohort based on the diagnosis code from Table 1A that appears in the greatest number of episodes of care during the study period, either primary or secondary.

To be eligible for the co-occurring disorders indicators clients must be assigned to one of the four mental health cohorts (e.g., MDD, PTSD, schizophrenia, bipolar disorder) and the SUD cohort. If a patient was assigned to one of these cohorts in FY06 and another in FY07, they will still qualify for the co-occurring disorders cohort.

TABLE 1A. COHORT DIAGNOSES

Cohort ICD-9 Code Description

295.0 Schizophrenia, simple type 295.1 Schizophrenia, disorganized type 295.2 Schizophrenia, catatonic type 295.3 Schizophrenia, paranoid type 295.4 Acute schizophrenic episode 295.5 Latent schizophrenia

295.6 Residual schizophrenia

295.7 Schizophrenia, schizo-affective type 295.8 Other specified types of schizophrenia Schizophrenia

295.9 Unspecified schizophrenia

PTSD 309.81 Prolonged posttraumatic stress disorder 296.0 Manic disorder, single episode

296.1 Manic disorder, recurrent episode 296.4 Bipolar affective disorder, manic 296.5 Bipolar affective disorder, depressed Bipolar

296.6

296.7 Bipolar affective disorder, mixed Bipolar affective disorder, unspecified 296.2 Major depressive disorder, single episode Major Depressive

Disorder 296.3 Major depressive disorder, recurrent episode 303.90-303.92 Other and unspecified alcohol dependence 304.00-304.02 Opioid type dependence

304.10-304.12 Barbiturate and similarly acting sedative or hypnotic dependence

304.20-304.22 Cocaine dependence 304.30-304.32 Cannabis dependence Substance Use

Disorder

304.40-304.42 Amphetamine and other psychostimulant dependence

(16)

14

Cohort ICD-9 Code Description

304.50-304.52 Hallucinogen dependence

304.60-304.62 Other and unspecified drug dependence 304.70-304.72 Combinations of opioid type with any other

304.80-304.82 Combinations of drug dependence excluding opioid type

304.90-304.92 Unspecified drug dependence 305.00-305.02 Alcohol abuse

305.20-305.22 Cannabis abuse 305.30-305.32 Hallucinogen abuse

305.40-305.42 Barbiturate and similarly acting sedative or hypnotic abuse

305.50-305.52 Opioid abuse 305.60-305.62 Cocaine abuse

305.70-305.72 Amphetamine or related acting sympathomimetic abuse

305.90-305.92 Other, mixed, or unspecified drug abuse Co-Occurring

Disorders

Cohort Diagnosis of SUD and MDD, PTSD, SUD, Bipolar or Schizophrenia

Defining Study-Relevant Treatment Encounters

Ib. Diagnostic codes used to define treatment encounters

The set of codes in Table 1B will be used to describe encounters or episodes occurring during treatment, to establish relevant treatment, and also to define the beginning of a new treatment episode. Codes in italics represent additional codes from those used to define the diagnostic cohort.

(17)

15

TABLE 1B. DIAGNOSES FOR DEFINING TREATMENT ENCOUNTERS

Diagnosis ICD-9 Code Description

295.0 Schizophrenia, simple type 295.1 Schizophrenia, disorganized type 295.2 Schizophrenia, catatonic type 295.3 Schizophrenia, paranoid type 295.4 Acute schizophrenic episode 295.5 Latent schizophrenia

295.6 Residual schizophrenia

295.7 Schizophrenia, schizo-affective type 295.8 Other specified types of schizophrenia Schizophrenia 295.9 298.9 295.40 298.8 293.xx 296.0 296.1 296.4 296.5 296.6 296.7 296.8x Unspecified schizophrenia

Psychosis disorder, not otherwise specified Schizophreniform Disorder

Brief psychotic disorder

Psychotic disorder due to a general medical condition

Manic disorder, single episode Manic disorder, recurrent episode Bipolar affective disorder, manic Bipolar affective disorder, depressed Bipolar affective disorder, mixed Bipolar affective disorder, unspecified Manic-depressive psychoses, other and unspecified

PTSD 309.81 Prolonged posttraumatic stress disorder 296.0 Manic disorder, single episode

296.1 Manic disorder, recurrent episode 296.4 Bipolar affective disorder, manic 296.5 Bipolar affective disorder, depressed 296.6 Bipolar affective disorder, mixed

296.7 Bipolar affective disorder, unspecified

Bipolar 296.8x 295.0 295.1 295.2 295.3 295.4 295.5 295.6 295.7 295.8 295.9 298.9 295.40 298.8 293.xx

Manic-depressive psychoses, other and unspecified

Schizophrenia, simple type Schizophrenia, disorganized type Schizophrenia, catatonic type Schizophrenia, paranoid type Acute schizophrenic episode Latent schizophrenia

Residual schizophrenia

Schizophrenia, schizo-affective type Other specified types of schizophrenia Unspecified schizophrenia

Psychosis disorder, not otherwise specified Schizophreniform Disorder

Brief psychotic disorder

Psychotic disorder due to a (general medical condition)

(18)

16

Diagnosis ICD-9 Code Description

Major Depressive Disorder

296.2 Major depressive disorder, single episode 296.3 311 Major depressive disorder, recurrent episode Depressive Disorder NOS

300.4 Dysthymia

293.83 Mood Disorder due to Medical Condition

296.90 Mood Disorder NOS

309.1 Prolonged Depressive Reaction

Major Depressive Disorder

296.99 Other Specified Affective Disorders

292.9

303.90-303.92

Opioid related disorder NOS

Other and unspecified alcohol dependence 304.00-304.02 Opioid type dependence

304.10-304.12 Barbiturate and similarly acting sedative or hypnotic dependence

304.20-304.22 Cocaine dependence 304.30-304.32 Cannabis dependence

304.40-304.42 Amphetamine and other psychostimulant dependence

304.50-304.52 Hallucinogen dependence

304.60-304.62 Other and unspecified drug dependence 304.70-304.72 Combinations of opioid type with any other 304.80-304.82 Combinations of drug dependence excluding

opioid type

304.90-304.92 Unspecified drug dependence 305.00-305.02 Alcohol abuse

305.20-305.22 Cannabis abuse 305.30-305.32 Hallucinogen abuse

305.40-305.42 Barbiturate and similarly acting sedative or hypnotic abuse

305.50-305.52 Opioid abuse 305.60-305.62 Cocaine abuse

305.70-305.72 Amphetamine or related acting sympathomimetic abuse

Substance Use Disorder

(19)

17

Defining New Treatment Episodes

IIa. Bipolar Disorder

The new treatment episode for bipolar disorder is defined as:

 A recent, diagnosis-related admission1 or transfer to an inpatient/residential mental health bed,

OR

 An outpatient encounter where bipolar disorder (Table 1B) is the primary diagnosis following a break in care.

Break in care is defined as:

o NO bipolar-related medications for 5 or more months

AND

o NO outpatient encounters where bipolar disorder is either the primary or the secondary diagnosis for 5 or more months.2

Patient cohorts: All patients in the bipolar disorder cohort (Table 1A).

Definitions:

• Bipolar Disorder Encounter: ICD-9 codes including 296.0, 296.1, 296.4, 296.5, 296.6, 296.7, 296.8x, 295.0, 295.1, 295.2, 295.3, 295.4, 295.5, 295.6, 295.7, 295.8, 295.9, 298.9, 295.40, 298.8, 293.xx

• Bipolar Medication: Any medications for which a prescription was filled. These include drugs from the following VA Drug Class Codes found in VHA Pharmacy Prescription Data:

o CN400, Anticonvulsants o CN600, Antidepressants

o CN601, Tricyclic Antidepressants

o CN602, Monamine Oxidase Inhibitor Antidepressants o CN609, Antidepressants, Other

o CN700, Antipsychotics

o CN701, Phenothiazine/Related Antipsychotics

1_{The new episode of care begins on the date of admission or transfer; however, the discharge}

diagnosis will be used for purposes of describing the admission/transfer.

2_{A break in care is defined as 5 or more months without condition-related medications or}

condition-related outpatient encounters. The definition of a break in care is fairly restrictive in order to address concerns that poor medication compliance could include patients who have a 90-day prescription and are still taking medications but inconsistently.

(20)

18

o CN709, Antipsychotics, Other o CN750, Lithium Salts

• Inpatient admission3 where any psychiatric diagnosis is the primary diagnosis (ICD-9 codes: 2(ICD-90.xx-31(ICD-9.xx) and, if the primary diagnosis is not a diagnosis in Table 1B, at least one secondary diagnosis comes from Table 1B.

• If it is impossible to determine which diagnosis for an outpatient encounter is the primary diagnosis, then a diagnosis in Table 1B must be listed as one of the diagnoses for the encounter.4

Instructions:

The start of the new treatment episode for bipolar disorder will be defined by:

1) The admission date or transfer date for any inpatient hospitalization as defined above.

OR

2) An outpatient encounter where bipolar disorder (Table 1B) is the primary diagnosis following a clean period of five or more months (based on a 90-day prescription) for which there is:

 NO prescription filled for selected medications,

AND (in the same time period of five or more months)

 NO outpatient encounter in any clinic where bipolar disorder is the primary or secondary diagnosis.

The first visit after the clean period in which bipolar disorder is the primary diagnosis will indicate the start date for the new treatment episode.

IIb. Schizophrenia

The new treatment episode for schizophrenia is defined as:

 A recent, diagnosis-related admission or transfer to an inpatient/residential mental health bed,

OR

 An outpatient encounter where schizophrenia (Table 1B) is the primary diagnosis following a break in care.

3_{Defining the NTE based on inpatient discharges was modified such that the primary diagnosis}

must be any psychiatric diagnosis (210.xx-319.xx) and, if the primary diagnosis was not one of those in Table 1B, an added requirement is that a diagnosis from Table 1B must be listed as a secondary diagnosis.

4_{Definition for how an outpatient encounter triggers a NTE was modified to be made consistent}

with the practicalities of the data being extracted from medical records by WVMI. It is not always possible to determine which of the diagnoses listed for an outpatient encounter is the primary diagnosis based on the clinical notes. In these cases, a diagnosis from Table 1B must be listed as one of the diagnoses for the encounter.

(21)

19

o NO schizophrenia-related medications for 5 or more months

AND

o NO outpatient encounters where schizophrenia disorder is either the primary or the secondary diagnosis for 5 or more months.5

Patient cohorts: All patients in the schizophrenia disorder cohort (Table 1A).

Definitions:

• Schizophrenia Encounter: ICD-9 codes including 295.x, 298.9, 295.40, 298.8, 293.xx, 296.0, 296.1, 296.4, 296.5, 296.6, 296.7, 296.8x.

• Schizophrenia Medication: Any medications for which a prescription was filled. These include drugs from the following VA Drug Class Codes found in VHA Pharmacy Prescription Data:

o CN701, Phenothiazine/Related Antipsychotics o CN709, Antipsychotics, Other

• Inpatient/residential mental health admission where any psychiatric diagnosis is the primary diagnosis (ICD-9 codes: 290.xx-319.xx) and, if the primary diagnosis is not a diagnosis in Table 1B, at least one secondary diagnosis comes from Table 1B.

Instructions:

The start of the new treatment episode for schizophrenia will be defined by:

OR

2) An outpatient encounter where schizophrenia (Table 1B) is the primary diagnosis following clean period of five or more months (based on a 90-day prescription) for which there is:

AND (in the same time period of five or more months)

 NO outpatient encounter in any clinic where schizophrenia is the primary or secondary diagnosis.

The first visit after the clean period in which schizophrenia is the primary diagnosis will indicate the start date for the new treatment episode.

condition-related outpatient encounters. The definition of a break in care is a fairly restrictive in order to address concerns that poor medication compliance could include patients who have a 90-day prescription and are still taking medications but inconsistently.

(22)

20

IIc. Major Depressive Disorder (MDD)

The new treatment episode for MDD is defined as:

OR

 An outpatient encounter where MDD (Table 1B) is the primary diagnosis following a break in care.

o NO MDD-related medications for 5 or more months

AND

o NO encounters where MDD disorder is either the primary or the secondary diagnosis for 5 or more months.6

Patient cohorts: All patients in the MDD cohort (Table 1A).

Definitions:

• MDD Encounter: ICD-9 codes including 293.83, 296.2, 296.3, 296.90, 296.99, 300.4, 309.1, 311, 300.4, 293.83, 296.90, 309.1, 296.99.

• MDD Medication: Any medications for which a prescription was filled. These include drugs from the following VA Drug Class Codes found in VHA Pharmacy Prescription Data:

o CN600, Antidepressants

o CN750, Lithium Salts

Instructions:

The start of the new treatment episode for MDD will be defined by:

(23)

21

OR

2) An outpatient encounter where MDD (Table 1B) is the primary diagnosis following clean period of five or more months (based on a 90-day prescription) for which there is:

AND (in the same time period of five or more months)

 NO outpatient encounter in any clinic where MDD is the primary or secondary diagnosis.

The first visit after the clean period in which MDD is the primary diagnosis will indicate the start date for the new treatment episode.

IId. Post-Traumatic Stress Disorder (PTSD)

The new treatment episode for PTSD is defined as:

OR

 An outpatient encounter where PTSD is the primary diagnosis following a break in care.

o NO PTSD-related medications for 5 or more months

AND

o NO outpatient encounters where PTSD is either the primary or the secondary diagnosis for 5 or more months.7

Patient cohorts: All patients in the PTSD cohort (Table 1A).

Definitions:

• PTSD Encounter: ICD-9 codes including 309.81.

condition-related outpatient encounters. The definition of a break in care is a fairly restrictive in order to address concerns that poor medication compliance could include patients who have a 90-day prescription and are still taking medications but inconsistently. Initially, a break in care was defined using a 12-month timeframe; we may modify this timeframe pending a discussion with PTSD experts.

(24)

22

• PTSD Medication: Any medications for which a prescription was filled. These include drugs from the following VA Drug Class Codes or NDC codes found in VHA Pharmacy Prescription Data:

o CN600, Antidepressants

o Prazosin (see Appendix A for associated NDC codes)

Instructions:

The start of the new treatment episode for PTSD will be defined by:

OR

2) An outpatient encounter where PTSD is the primary diagnosis following clean period of 5 or more months for which there is:

AND (in the same time period of twelve or more months)

 NO outpatient encounter in any clinic where PTSD is the primary or secondary diagnosis.

The first visit after the clean period in which PTSD is the primary diagnosis will indicate the start date for the new treatment episode.

IIe. Substance Use Disorder (SUD)

The new treatment episode for SUD is defined as:

OR

 An outpatient encounter where SUD is the primary diagnosis following a break in care.

(25)

23

o NO SUD-related medications for 5 or more months

AND

o NO outpatient encounters where SUD is either the primary or the secondary diagnosis for 5 or more months.8

Patient cohorts: All patients in the SUD cohort (Table 1A).

Definitions:

• SUD Encounter: ICD-9 codes including 303.9, 304.0x-304.9x, 305.0x-305.9x (where ‘x’ equals 0, 1, or 2).

• SUD-Related Medication: Any medications for which a prescription was filled. These include drugs from the following VA Drug Class Codes or NDC codes found in VHA Pharmacy Prescription Data or from stop codes:

o AD100, Alcohol Deterrents (for alcohol abuse/dependence, 303.9, 305.0) o Naltrexone (for alcohol abuse/dependence, 303.9, 305.0): see Appendix

A for the associated NDC codes

o Methadone (for opiate addiction, 304.0, 304.7, 305.5): defined by stop codes for opioid substitution (stop code: 523)

o Buprenorphine (for opiate abuse/dependence, 304.0, 304.7, 305.5): see Appendix A for the associated NDC codes

Instructions:

The start of the new treatment episode for SUD will be defined by:

1) The admission or transfer date for any inpatient hospitalization as defined above.

OR

2) An outpatient encounter where SUD is the primary diagnosis following a clean period of five or more months (based on a 90-day prescription) for which there is:

AND (in the same time period of five or more months)

 NO outpatient encounter in any clinic where SUD is the primary or secondary diagnosis.

(26)

24

The first visit after the clean period in which SUD is the primary diagnosis will indicate the start date for the new treatment episode.

IIf. Care for Co-Occurring Disorders

To be eligible for the co-occurring disorders indicators, clients must be assigned to one of the four mental health cohorts (e.g., MDD, PTSD, schizophrenia, bipolar disorder) and the SUD cohort.

Defining an Index Visit for those with co-occurring disorders: For relevant co-occurring

disorders performance indicators, we will identify the start date for evaluation based on the identification of an index visit for veterans identified as having a co-occurring

disorder as defined above. The index visit is the first encounter during the study period that meets one of the following criteria:

1.

The first encounter in which both the veteran’s mental health condition and SUD are present among the ICD-9 codes listed on the administrative file (either the mental health condition or SUD should be primary).

2.

If the first encounter of the study period is for a mental health encounter (primary diagnosis only – Table 1B) without mention of an SUD diagnosis, the previous 6 month period and the 6 month period following that encounter are scanned for evidence of an encounter for SUD (primary or secondary diagnosis – Table 1B). If the SUD encounter precedes the mental health encounter by less than six months, the mental health encounter is labeled the index visit for purposes of evaluating the performance indicators. If there is no prior SUD encounter but there is an SUD encounter within the six months following the mental health encounter, that SUD encounter is labeled the index visit for purposes of evaluating the performance indicators.

3.

Alternatively, if the first encounter of the study period is an encounter for SUD (primary diagnosis only – Table 1B) without mention of a mental health diagnosis, the six months prior to and following that encounter are scanned to find evidence of a mental health encounter (primary or secondary diagnosis – Table 1B). If the mental health encounter precedes the SUD encounter by less than six months, the SUD encounter is labeled the index visit for purposes of evaluating the performance indicators. If the mental health encounter follows the SUD encounter by less than six months, the mental health encounter is labeled the index visit for purposes of evaluating the performance indicators.

4.

If the first encounter of the study period of either type (mental health or SUD without mention of the other diagnosis) is not preceded by or followed by another relevant encounter within six months, the encounter stream is followed through the study period until the above criteria are met.

o Note: It may be possible for someone to be classified as having a co-occurring disorder and not have an index visit during the study period.

(27)

25

Additional Concepts IIIa. Specialty Mental Health

Includes diagnosis-related (either primary or secondary using Table 1B) visits to any specialty mental health provider and includes both psychiatry and substance abuse care. The following codes from the Medical SAS Outpatient Datasets (PROV1-PROV10) define specialty mental health care:

• 010000, 010100, 010200-010206, 010300-010302, 010400-010421, 070101, 070102, 070807, 070901, 070947, 070953, 100324-100330, 110100, 114100, 114800-114810, 114900, 115000, 118324, 181001, 182901-182913

IIIb. Licensed Mental Health Provider

Includes diagnosis-related (either primary or secondary using Table 1B) visits to

physicians (MD, DO), physician assistants, nurse practitioners/clinical nurse specialists and psychologists/psychoanalysts with a mental health specialty. The following codes from the Medical SAS Outpatient Datasets (PROV1-PROV10) define licensed mental health providers: • 110100, 114100, 114800-114810, 115000, 118324, 181001, 182901-182913, MD/DO • 010300-010421, Psychologists/Psychoanalysts • 100324-100330, PA/CNS • 100616, NP

• 010100, Clinical Social Worker

IIIc. Licensed Mental Health Prescribing Provider

Includes diagnosis-related (either primary or secondary using Table 1B) visits to physicians (MD, DO), physician assistants, and nurse practitioners/clinical nurse

specialists with a mental health specialty and can prescribe medications. The following codes from the Medical SAS Outpatient Datasets (PROV1-PROV10) define licensed mental health prescribing providers:

• 114100, 114800-114810, 115000, 118324, 182901-182913, MD/DO • 100324-100330, PA/CNS

(28)

26

IIId. Licensed Prescribing Provider

Includes diagnosis-related (either primary or secondary using Table 1B) visits to physicians (MD, DO), physician assistants, and nurse practitioners/clinical nurse specialists with any specialty (not limited to mental health) and can prescribe

medications. The following codes from the Medical SAS Outpatient Datasets (PROV1-PROV10) define licensed prescribing providers:

• 110000, 110100, 110200, 110300, 110301, 110400-110404, 110500, 110600, 110700-110703, 110800-110804, 110900-110902, 111000, 111100, 111200, 111300, 111400, 111500-111521, 111600, 111610, 111700, 111800, 111900-111906, 112100, 112101, 112200, 112300, 112400, 112500, 112600, 112601, 112800, 112900-112904, 113000, 113001, 113100, 113001, 113200, 113300-113302, 113400, 113500, 113600, 113700-113716, 113800, 113900, 114000-114002, 114100, 114200-114216, 114300, 114400, 114500-114504, 114600-114608, 114700, 114800-114810, 114900, 115000, 115100, 115200, 115300-115309, 115400, 115500, 115600, 115700-115705, 115800, 115900, 116000, 116100, 116200, 116300, 116400, 116500, 116600, 116700, 116800, 116801, 116900, 117000, 117100, 117200, 117300, 117400-117403, 117500, 117600, 117700, 117800, 117900, 118000, 118100, 118200, 118201, 118301-118342, 118434, 160100, 180100-180102, 180200-180203, 180300, 180400, 180500-180505, 180600-180605, 180700-180705, 180800, 180900, 181000-181021, 181100, 181200, 181301-181305, 181400, 181500, 181600, 181700-181703, 181800-181806, 181900, 182000, 182100-182106, 182200-182206, 182301, 182302, 182401-182414, 182500-182518, 182600-182604, 182700-182702, 182801-182807, 182901-182913, 183001-183010, 183100-183107, 183200, 183300, 183400, MD/DO • 100000, 100100, 100200, 100300-100335, PA/CNS • 100600-100618, NP

IIIe. Specialty SUD Care

Defined as one or more SUD outpatient visits in the first 30 days following the start of an new treatment episode. Appropriate visits include stop codes for SUD treatment found in the Medical SAS Outpatient Dataset:

• 513: Substance Abuse – Individual • 514: Substance Abuse – Home Visit

• 519: Substance Use Disorder/PTSD Teams • 523: Opioid Substitution

• 545: Telephone/Substance Abuse

• 547: Intensive Substance Abuse Treatment • 560: Substance Abuse - Group

IIIf. Psychotherapy

Any diagnosis-related (either primary or secondary using Table 1B) clinic encounters for which the following CPT codes are present:

(29)

27

• 90806, 90807, 90808, 90809 Office or Other Outpatient Facility, Insight Oriented, Behavior Modifying and/or Supportive Psychotherapy, excluding psychotherapy with medical evaluation and management services less than 30 minutes

• 90812, 90813, 90814, 90815, Office or Other Outpatient Facility, Interactive

Psychotherapy (Note: these codes most likely apply to psychotherapy with children but will be retained for evaluation purposes in case they may be used with adult patients), excluding psychotherapy with medical evaluation and management services less than 30 minutes

• 90818, 90819, 90821, 90822, Inpatient Hospital, Partial Hospital or Residential Treatment Facility, excluding psychotherapy with medical evaluation and management services less than 30 minutes

• 90826, 90827, 90828, 90829, Inpatient Hospital, Partial Hospital or Residential Treatment Facility, Interactive Psychotherapy (Note: these codes most likely apply to psychotherapy with children but will be retained for evaluation purposes in case they may be used with adult patients), excluding psychotherapy with medical evaluation and management services less than 30 minutes

• 90845, Psychoanalysis

• 90853, Group Psychotherapy (other than of a multiple-family group)

• 90857, Interactive Group Psychotherapy (Note: these codes most likely apply to psychotherapy with children but will be retained for evaluation purposes in case they may be used with adult patients.)

(30)

28

(31)

29

SCHIZOPHRENIA

Performance Measure Technical Documentation Module: Schizophrenia

Indicator Statement: Proportion of selected schizophrenia patients with appropriate

short-term utilization of antipsychotic medications

Indicator number: A

Executive Summary: This indicator is based on a quality measure cited by the Center

for Quality Assessment and Improvement in Mental Health (CQAIMH) and developed by Lehman et al (1998, updated 2004) for the AHRQ-funded Schizophrenia PORT program. CQAIMH states that “Antipsychotic medications have been shown to be efficacious in the treatment of acute psychotic exacerbations of schizophrenia and in reducing the likelihood of relapse.” According to the American Psychiatric Association Clinical Practice Guidelines, “The evidence supporting the effectiveness of first-generation antipsychotic medications in reducing psychotic symptoms in acute schizophrenia comes from studies carried out in the 1960s as well as numerous subsequent clinical trials,” and “pharmacological treatments should be initiated as soon as is clinically feasible, because acute psychotic exacerbations are associated with emotional distress, disruption to the patient’s life, and a substantial risk of [violent] behavior.” Numerous studies have shown that antipsychotic treatments are effective in the acute phase of schizophrenia. The strength of the evidence led the researchers involved in the AHRQ-funded Schizophrenia Patient Outcomes Research Team (PORT) to recommend this practice which they rated as having ‘good research-based evidence’ (Lehman et al, 1998, updated 2004). The VA Clinical Practice Guidelines also support the use of antipsychotic drugs in schizophrenia treatment.

While antipsychotic medications have been clinically shown to be effective in the treatment of schizophrenia, we cannot account for medication refusals or

contraindications in this analysis. This indicator addresses the following IOM domain: effectiveness

References:

Anthony F. Lehman, Donald M. Steinwachs, and The Co-Investigators of the PORT Project, “Translating Research Into Practice: The Schizophrenia Patient Outcomes Research Team (PORT): Updated Treatment Recommendations 2003”,

Schizophrenia Bulletin 30: 2 (2004)

Center for Quality Assessment and Improvement in Health Care, Quality indicators published online at www.cqaimh.org. Accessed in October 2007

Management of Psychoses. Washington, DC: VA/DoD Evidence-Based Clinical Practice Guideline Working Group, Veterans Health Administration, Department of Veterans Affairs, and Health Affairs, Department of Defense, May 2004. Office of Quality and Performance publication 10Q-CPG/PSY-04

Practice Guideline for the Treatment of Patients with Schizophrenia, Second Edition (February 2004); American Psychiatric Association; Am J Psychiatry

Numerator:

a) Patients prescribed an antipsychotic for 12 weeks following the start of a new treatment episode

(32)

30

b) Patients prescribed an antipsychotic for less than 12 weeks following the start of a new treatment episode

c) Patients with no filled prescription for an antipsychotic during the 12 weeks following the start of a new treatment episode

Denominator: All patients with schizophrenia in a new treatment episode Patient cohorts: Patients with schizophrenia diagnosis.

Definitions:

• New Treatment Episode: See the Key Definitions Document

o If the new treatment episode begins in the inpatient setting, the start of a new treatment episode is defined as the admission date

o If the new treatment episode begins in the outpatient setting, the start of the new treatment episode is defined as the first diagnosis-related (primary diagnosis only using Table 1B in the Key Definitions Document) outpatient encounter following the 5 month clean period

• 12 weeks of an antipsychotic: Defined as 60 or more days supplied of an antipsychotic in the 12 weeks following the date of the first filled prescription subsequent to the start of a new treatment episode

o If the new treatment episode begins in the inpatient setting, the patient is considered to be fully compliant with their medications while in the hospital and so the number of days supplied should include the length of stay in the hospital that begins the new treatment episode (using the variable ‘LS’ from the Medical SAS Inpatient Dataset)

o The days of medication should be sequential and should not include simultaneous use of two or more qualifying medications over a shorter period of time.

• Less than 12 weeks of an antipsychotic: Defined as at least one filled

prescription but with less than 60 days supplied in the 12 weeks following the date of the first filled prescription subsequent to the start of a new treatment episode

• No filled prescription: Defined as no filled prescriptions for any antipsychotic within 12 weeks of the start of the new treatment episode

• Antipsychotic medications: One or more prescriptions filled for a patient using the following drug class and NDC codes:

Strength of Evidence: Grade III – This is the level of evidence for the indicator as

specified, as we cannot account for medication refusals, poor treatment adherence, and other factors beyond the control of the prescriber in this analysis. However, the

evidence cited in the executive summary has a strength of evidence of Grade I.

Feasibility/Data Collection Issues:

• We do not include patients receiving depot antipsychotics in the numerator of this indicator.

(33)

31

Updates:

• The days of medication should be sequential and should not include

simultaneous use of two or more qualifying medications over a shorter period of time.

(34)

32

Performance Measure Technical Documentation Module: Schizophrenia

Indicator Statement: Proportion of schizophrenia patients with long-term utilization of

antipsychotic medications

Indicator number: B

for Quality Assessment and Improvement in Mental Health (CQAIMH), which states that “Antipsychotic medications have been shown to be efficacious in the treatment of acute psychotic exacerbations of schizophrenia and in reducing the likelihood of relapse.” Likewise, for outpatients, “Controlled trials have shown that patients who receive

antipsychotic medication for 1 year after an acute psychiatric episode experience a lower likelihood of relapse compared to patients treated with a placebo.” The evidence for this indicator was generated by Lehman et. al. (1998, updated 2004) for the AHRQ-funded Schizophrenia PORT program who report that “on average, persons on maintenance therapy experienced symptom relapse over a follow-up year at a rate of about 20 to 25 percent compared with about 55 percent for those on placebo. The value of

maintenance therapy beyond the first year has not been studied extensively.” According to the American Psychiatric Association Clinical Practice Guidelines, “Antipsychotic medications substantially reduce the risk of relapse in the stable phase of illness and are strongly recommended.” The VA Clinical Practice Guidelines also support the use of antipsychotic drugs in schizophrenia treatment.

While antipsychotic medications have been clinically shown to be effective in the treatment of schizophrenia, we cannot account for medication refusals or

contraindications in this analysis. This indicator addresses the following IOM domain: effectiveness

References:

Practice Guideline for the Treatment of Patients with Schizophrenia, Second Edition (February 2004); American Psychiatric Association; Am J Psychiatry

Anthony F. Lehman, Donald M. Steinwachs, and The Co-Investigators of the PORT Project, “Translating Research Into Practice: The Schizophrenia Patient Outcomes Research Team (PORT): Updated Treatment Recommendations 2003”,

Schizophrenia Bulletin 30: 2 (2004)

Management of Psychoses. Washington, DC: VA/DoD Evidence-Based Clinical Practice Guideline Working Group, Veterans Health Administration, Department of Veterans Affairs, and Health Affairs, Department of Defense, May 2004. Office of Quality and Performance publication 10Q-CPG/PSY-04

Numerator: Those individuals who received an antipsychotic medication for the

following periods of time:

a) Patients with 12 months supply of an antipsychotic medication during the study period

b) Patients with at least one filled prescription of an antipsychotic during the study period

(35)

33

Denominator: All patients with a schizophrenia diagnosis Patient cohorts: Patients with a diagnosis of Schizophrenia Definitions:

• 12 months of continuous antipsychotic medication: Defined as at least 300 days of the medication supplied during a 12-month period following the date of the first filled prescription during the study period.

o The days of medication should be sequential and should not include simultaneous use of two or more qualifying medications over a shorter period of time.

• Any use of an antipsychotic: Defined as at least one filled prescription for an antipsychotic during the study period

• No antipsychotic: Defined as no filled prescriptions for an antipsychotic during the study period

• Antipsychotic medications: Defined as one or more prescriptions using the following VA Drug Class Codes found in VHA Pharmacy Prescription Data:

Strength of Evidence: Grade III – This is the level of evidence for the indicator as

specified, as we cannot account for medication contraindications in this analysis. However, the evidence cited in the executive summary has a strength of evidence of Grade I.

Feasibility/Data Collection Issues:

• We do not include patients receiving depot antipsychotics in the numerator of this indicator.

Updates:

• The days of medication should be sequential and should not include

simultaneous use of two or more qualifying medications over a shorter period of time.

(36)

34

Performance Measure Technical Documentation Module: Schizophrenia

Indicator Statement: Proportion of selected schizophrenia patients with antipsychotic

polypharmacy utilization

Indicator number: C

for Quality Assessment and Improvement in Mental Health (CQAIMH) based on a study conducted by Poplin et al (1998), and a recent report by Kreyenbuhl et. al. (2007). CQAIMH states, “prescription of multiple medications requires caution related to potential drug interactions and side effects.” Kreyenbuhl et. al. support this assertion, noting that “treatment with multiple antipsychotic agents may present risks to patients, such as increased adverse effects and drug interactions, decreased treatment

adherence, increased costs for patients and health care systems, and a possible increased risk of mortality.” As reported by Ganguly et al (2004), “antipsychotic

polypharmacy is widely prevalent, is prescribed for long durations, and is an increasing phenomenon among (publicly insured) schizophrenia patients, indicating a significant discrepancy with treatment guidelines.” The American Psychiatric Association Clinical Practice Guidelines do not support the use of multiple antipsychotic medications. There is however almost no rigorous research on the effects of antipsychotic polypharmacy: most of the evidence has been generated by case reports or open, uncontrolled, nonrandomized studies (Waddington et. al. (1998); Centorrino et al (2004). Furthermore, there is some documentation that antipsychotic polypharmacy does not increase side effects (Ganesan 2008), although there is no methodologically sound empirical evidence that the practice is effective or free of safety concerns. For these reasons, this indicator will be used descriptively as an indication of practice variation, rather than as a benchmark of quality of care. This indicator addresses the following IOM domain: safety and effectiveness

References:

Am J Psychiatry, Apr 2004; 161: 700 - 706. Practice Guideline for the Treatment of Patients with Schizophrenia, Second Edition (February 2004); American Psychiatric Association; Am J Psychiatry

Centorrino Franca, Jessica L. Goren, John Hennen, Paola Salvatore, James P. Kelleher, and Ross J. Baldessarini. Multiple Versus Single Antipsychotic Agents for Hospitalized Psychiatric Patients: Case-Control Study of Risks Versus Benefits. Ganesan, Soma. Antipsychotic polypharmacy does not increase the risk for side

effects. Schizophrenia Research 98 (2008) 323–324

Ganguly R., Kotzan J.A., Miller L.S., Kennedy K., Martin B.C. (2004). Prevalence, trends, and factors associated with Antipsychotic polypharmacy among medicaid-eligible schizophrenia patients, 1998-2000. J Clin Psychiatry, 65:1377-88.

Kreyenbuhl, Julie A., Valenstein, Marcia, McCarthy, John F., Ganoczy, Dara, Blow, Frederic C. “Long-Term Antipsychotic Polypharmacy in the VA Health System: Patient Characteristics and Treatment Patterns.” Psychiatric Services 2007 58: 489-495.

Management of Psychoses. Washington, DC: VA/DoD Evidence-Based Clinical Practice Guideline Working Group, Veterans Health Administration, Department of

(37)

WORKING P A P E R. Veterans Health Administration Mental Health Program Evaluation Technical Manual

Veterans Health

Preface

TABLE OF CONTENTS

PART I: Key Definitions Document... 12

PART II: ADMINISTRATIVE DATA INDICATORS ... 28

Part III: Medical Records Review Indicators ... 117

SCHIZOPHRENIA ... 119

Discussion

INTRODUCTION

Strength of evidence:

Study Period:

Defining Study-Relevant Populations

TABLE 1A. COHORT DIAGNOSES

Cohort ICD-9 Code Description

Cohort ICD-9 Code Description

Defining Study-Relevant Treatment Encounters

TABLE 1B. DIAGNOSES FOR DEFINING TREATMENT ENCOUNTERS

Diagnosis ICD-9 Code Description

Diagnosis ICD-9 Code Description

Defining New Treatment Episodes

Patient cohorts: All patients in the bipolar disorder cohort (Table 1A).

Instructions:

AND (in the same time period of five or more months)

Patient cohorts: All patients in the schizophrenia disorder cohort (Table 1A).

Instructions:

AND (in the same time period of five or more months)

IIc. Major Depressive Disorder (MDD)

Patient cohorts: All patients in the MDD cohort (Table 1A).

Instructions:

AND (in the same time period of five or more months)

Patient cohorts: All patients in the PTSD cohort (Table 1A).

Instructions:

AND (in the same time period of twelve or more months)

Patient cohorts: All patients in the SUD cohort (Table 1A).

Instructions:

AND (in the same time period of five or more months)

Additional Concepts IIIa. Specialty Mental Health

IIIc. Licensed Mental Health Prescribing Provider

IIId. Licensed Prescribing Provider

IIIf. Psychotherapy

SCHIZOPHRENIA

Performance Measure Technical Documentation Module: Schizophrenia

References:

Numerator:

Denominator: All patients with schizophrenia in a new treatment episode Patient cohorts: Patients with schizophrenia diagnosis.

Strength of Evidence: Grade III – This is the level of evidence for the indicator as

Updates:

Indicator number: B

References:

Numerator: Those individuals who received an antipsychotic medication for the

Strength of Evidence: Grade III – This is the level of evidence for the indicator as

Performance Measure Technical Documentation Module: Schizophrenia

References: